on
Blog Post 5
##
## Call:
## lm(formula = adVoteShare$repAdPercent ~ adVoteShare$RepVotesMajorPercent)
##
## Residuals:
## Min 1Q Median 3Q Max
## -62.427 -12.435 -0.428 11.319 78.426
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -15.51708 4.49545 -3.452 0.000592 ***
## adVoteShare$RepVotesMajorPercent 1.21176 0.08524 14.216 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 19.99 on 669 degrees of freedom
## (21 observations deleted due to missingness)
## Multiple R-squared: 0.232, Adjusted R-squared: 0.2308
## F-statistic: 202.1 on 1 and 669 DF, p-value: < 2.2e-16
Gross Rating Points (GRP) is a measure of how many people are watching television programming when an ad is aired. This is really important because in arguably this is more meaningful than fundraising or ad spending is. Even if assume that ad spending is effective in persuading some voters, ad spending on its own is only impactful if people actually see the ads. We calculate this score by thinking about the percent of the ad’s media market that will see the ad. For example, in the article, Huber et al say that a GRP of 50 would mean half of the households in a media market saw a given ad. With this data, we can do things like see how a candidate’s advantage in spending (spending peaks) might correspond with higher GRPs. With this in mind, I would think that in order to approximate advertising effects at the national level like a Huber type study or even at a district level, we would want to use the lessons from the readings and consider not just the amount of ads being aired or bought but also the amount of viewership of the ads through things like the GRP. Especially for district level races where a lot of the races might not have large numbers of ads being aired, looking at the viewership of ads might allow us to tell if advertising truly mattered. For example, if we saw a correlation between a candidate’s peak in support and the amount of advertising being spent, but then also saw that the ads had very low viewership or had low GRPS, we may want to question whether or not the ads were casuing the peak in support. Maybe more ads are bought when outside factors are causing high levels of support and the campaign has more money. While it would be difficult for us to hone this conclusion down entirely, having access to information like GRPs and potentially seeing that some ads actually weren’t even being viewed at high levels might allow us to conclude that the advertising was not the important factor at play.